Substance abusing delinquents are on a developmental trajectory to become deeply enmeshed in the adult criminal justice system, including high incarceration rates, associated criminal victimization, and large confinement costs. The costs associated with juvenile delinquency and substance use are staggering. Nationwide, the average daily cost is approximately $240 million, with annual costs exceeding $3 billion for detention and $5.7 billion for residential placement. Many evidence-based treatments (EBT) have shown promise in diverting delinquents from a costly trajectory into the adult criminal justice system. Over the past decade, the number of confined youths dropped by 55% producing both huge cost savings to states and the promise of better outcomes for youth and families. These reductions in detention occurred by widening the net (i.e., both more high risk and low risk juveniles) of referrals to community interventions. Unfortunately, little researc has identified how youth from across the risk spectrum respond differentially to treatment or the persistence of treatment effects. This research is critical in tailoring referral processes to more efficiently address specific needs of youth. The main objective of this secondary data analysis project is to evaluate a large scale, statewide implementation of the Redirection Project in Florida's Department of Juvenile Justice, an initiative aimed at redirecting youth from residential commitment to evidence-based, family-focused community interventions, such as Functional Family Therapy (FFT). We will follow juveniles' criminal behavior for five years (2006 to 2012) after their arrest and enrollment in Redirection FFT services. Tracking these youth into young adulthood, will provide not only an estimate of the short-term effectiveness and cost-effectiveness of FFT, but also determine if the effects are sustained into emerging adulthood when alcohol, marijuana, and opioid use and arrests are at their peak. The proposed evaluation capitalizes on the naturalistic training and implementation of FFT to 5,100 families across 11 judicial circuit courts in Florida. The first aim is to compare outcomes from the FFT sample (matched n ~ 4,100) to a one-to-one propensity matched sample receiving general probation services (PSAU matched n ~ 4,100). A second comparison is between FFT and a propensity matched sample receiving residential treatment (RT matched n ~ 4,100). Researchers are increasingly using Propensity Score Estimates (PSE) in quasi-experimental designs to create matched samples to permit comparisons of treatment and control conditions when RCTs are not feasible. A second objective is to examine the extent to which quality of FFT implementation is associated with recidivism, including the extent to which pre-treatment youth and family variables are associated with implementation quality and outcomes. This will help identify critical predictors of negative outcomes and to identify recalcitrant sub-groups of youth that might benefit from additional treatment development or enhanced training of therapists. A third aim is to assess the extent to which individual and community level risk factors and characteristics predict 5 year recidivism rates.